test_runner: remove runtime experimental warning

[cjihrig]

This commit removes the experimental warning that is emitted when the test runner is used. The test runner feature is still considered experimental, but this change makes its output easier to read.

[tniessen]

±0 as long as it remains experimental. (If it's just in the docs, it's easy to ignore, so if we later break the test runner, it might make users who rely on it unhappy.)

[cjihrig]

A couple notes:

• It would definitely still be considered experimental. I also think we should try to avoid breaking changes to the test runner at this point.
• Currently, the warning is emitted for every test file. When using the CLI runner with many files, this can get noisy if there are failures across several files. Fortunately, this only shows up for tests with failures, but...
• There is ongoing work to add a TAP parser which would ideally surface user output to both stdout and stderr, which would make the current behavior noisy even if nothing fails.
• There is prior art for dropping the runtime warning when it interferes with the usability of the feature. I'm thinking of util.parseArgs() specifically, but there are probably others.

[nodejs-github-bot]

CI: https://ci.nodejs.org/job/node-test-pull-request/47011/

[nodejs-github-bot]

CI: https://ci.nodejs.org/job/node-test-pull-request/47033/

[nodejs-github-bot]

CI: https://ci.nodejs.org/job/node-test-pull-request/47042/

[nodejs-github-bot]

Landed in ad8442f